Last week’s blog introduced a new white paper, The Role of Preferred Shares in Your Portfolio, coauthored with my colleague Raymond Kerzérho, director of research at PWL Capital. That article looked at the reasons investors might consider adding Canadian preferred shares to a diversified portfolio: namely high yields relative to corporate bonds, tax-favoured dividend income, and low correlation with other asset classes.
Those are three tempting reasons to use preferred shares. But as Raymond and I explain in our paper, the overall risk-reward trade-off in this asset class is not particularly compelling. In my opinion, most balanced portfolios would likely be better off without preferred shares. For those who want to add them to the mix, we make the following recommendations:
Only use preferreds in non-registered accounts. While preferred shares may be a diversifier in any portfolio, their largest benefit is their tax-advantaged dividend income. If you need current income from a non-registered account, preferreds can be a good alternative to corporate bonds, which are often very tax-inefficient. While preferreds carry more risk than bonds issued by the same corporation, this tax advantage should provide adequate compensation.
We do not believe preferred shares are appropriate for tax-sheltered accounts such as RRSPs or TFSAs. Once you remove their tax advantage, preferred shares do not offer sufficient returns to justify their additional risks compared with high-quality bonds.
Keep your allocation small. There is no perfect rule to determine the maximum weight that should be allocated to preferred shares. But given the risks and limited rewards, we believe most investors should limit their exposure to about 5% or 10% of their overall portfolio. This might be pushed to 15% for investors who have a large allocation to fixed income (70% or more) in taxable accounts.
Diversify broadly. Diversification is always important, but it is paramount in an asset class that has little upside potential. If you hold a small number of preferred shares, there is little or no opportunity to recover from the default of one issue, since preferred shares deliver their value gradually over time, piling up small gains in the form of dividends. Credit losses are rare with preferred shares, but they are potentially devastating, so it is crucial to diversify across many issues.
Don’t invest in individual preferreds. Many investors attempt to diversify their preferred share holdings by selecting a number of individual issues. But we believe investors should avoid individual preferred shares for at least three reasons:
- It is difficult to get adequate diversification with individual preferreds. Holding 10 or 20 issues does not provide enough protection, and assembling a portfolio with a larger number of holdings is impractical for most investors.
- The complexity of preferred shares makes it extremely difficult to select individual securities. While it is tempting to look only at yield and credit quality, investors also need to have a complete understanding of the embedded options and other features in each issue, and few people have the skill to do so. This complexity makes selecting preferred shares challenging for retail investors, and even advisors.
- Individual preferred shares are notoriously illiquid. They frequently trade with wide bid-ask spreads, making them expensive and difficult for investors to buy and sell.
When advisors at PWL Capital use preferred shares in client portfolios, they typically use the BMO S&P/TSX Laddered Preferred Share (ZPR). This ETF has the benefit of relatively low cost (0.51% MER), broad diversification (over 140 issues), and much better liquidity than you would get from holding individual preferreds. It also includes only rate reset preferred shares, which carry less interest rate risk than perpetual preferreds (this idea is explained in the white paper).
In my next blog, I’ll explore a controversial idea: are preferred shares an asset class where active management can add value over an indexed approach?
If an investor is searching for yield, would high yield bonds (e.g. XHB, XHY) be a better alternative than preferred shares for a tax-sheltered account?
Hey Dan, what are your thoughts on XPF? Gives you U.S. and some European exposure and with a reasonable MER.
@Andrew: My first reaction is to ask whether it makes sense to search for yield rather than considering total return. That’s especially true in a tax-sheltered account, where presumably you are just reinvesting any income. In any case, high-yield bonds generally have an even worse risk-reward trade-off than preferreds. Might be a good topic for another white paper. In the meantime this may be of interest, even if it’s a bit out of date:
https://canadiancouchpotato.com/2010/10/04/high-yield-bonds-and-your-portfolio-part-1/
@Tim: The problem is that non-Canadian preferreds do not qualify for the dividend tax credit, which removes the primary advantage of the asset class. In a taxable account, high-yielding foreign securities with little or no expected capital gains are are very unattractive.
I guess, I still don’t get why anyone would want them. Perferreds seem like a financially engineered product that will always benefit the issuer more than the investor. History seems to bear this out.
For example if we compare ZPR and CBO, we see that they are both ladders and have similar interest rate risk exposure. However, CBO has less credit risk (being more senior credit), lower volatility and historically better performance.
Is there any history of preferreds ever performing well vs. simpler alternatives like bonds? All I see is weaker performance than short term corp. bonds and risks like stocks (-20% drop in 2008).
As for their tax advantage… I think the phrase that comes to mind is; a bad investment is still a bad investment after a tax break.
I think preferreds just appeal to naïve investors chasing yield rather than total return.
@Brian: As I’m sure you can tell from the paper, I share the skepticism about this asset class. In a tax-sheltered account it’s an easy decision to choose bonds over preferreds. But in a taxable account, the difference between ZPR and CBO (to use your example) would be very large: if you were to assume the same before-tax return, the after-tax return on ZPR would be significantly higher. Whether that’s enough to compensate for the additional risk is up to the investor to decide, but it’s at least worth a conversation.
Okay, good to hear.
One small point, since it’s inception ZPR is down -8% (total return), CBO is up about 7% in that same time period. So even if you were taxed double on CBO vs. ZPR, CBO would still be ahead. Maybe that’s an unfairly short time period but even when you look at CPD vs. XSB over CPD’s life, a similar thing happens because of the drop that CPD had in 2008. The common scenario seems to be that preferreds never win; even in a taxable account.
Even in a fully taxable account, it seems to me that a simple, low cost 20%/80% (or even 15%/85%) portfolio of a broad stock fund and a GIC ladder will always be better than preferreds. Yes, you will pay more taxes on the GIC income but you’ll gain it by having a little stocks. Interestingly, a 20%/80% portfolio is also much less volatile than CPD even thru 2008. If you wanted to get complicated you could use a dividend fund like CDZ to lower taxes. Additionally, you could do strategic tax harvesting. Alternate approaches could be to use HXT to completely differ taxes on the stock side. There are lots of options.
I guess, all I am saying is that if you think you want/need preferreds; there is probably always a better alternative.
I’m looking forward to the next post. Will you interview James Hymas? (I suspect the answer is “yes”, due to their illiquid nature; certainly it’s the only area in my portfolio where I employ active management.)
@Nathan: No, I did not interview James Hymas. :)
Early in 2014 I purchased CPD and ZPR (for averaging effect because I couldn’t figure out which was more appropriate from reading the CCP and other rational literature) in small equal portions in my non-registered portfolio. My rationale was that I needed retirement income, and my prior strategy of deriving the necessary dividend income from standard Canadian equity ETF bulged out the proportion of that asset to an unacceptable percentage; also the Bond-like component of preferred share ETFs, I reasoned, helped to fill out my perceived Bond asset allocation (the remainder I filled with BXF).
I struggled with not really understanding the risks and rewards, and the fact that the MER was rather higher than my other cheap plain vanilla components, and how exactly to compartmentalise the Equity equivalent and the Bond-like equivalent of these instruments to match my risk tolerance. Finally late in the year we rationalised our retirement finances and converted our RRSP to RRIF, providing a regular income stream. So I sold the preferreds off, and by that time HBB had come along, (or more correctly, my acceptance of HBB had finally come) solving my other problem of needing a bond component, but not being willing to pay the tax consequences of having regular Bond ETFs in a taxable account, allowing me to sell my BXF (as we don’t need further income) and to replace the whole lot, (that is, the bond asset allocation) with HBB.
So now you’re explaining the risk/reward situation that I didn’t really understand before and it’s quite interesting.
With this knowledge I briefly toyed with re-introducing ZPR in a small percentage for diversification (the correlation figures are encouraging), but in the long run, the protection it might give seems trivial compared to the extra hassle, and the somewhat higher MER than my average holding. We’ll wait for next week to see if any compelling previously un-considered reason to include ZPR pops up.
@Brian G: I considered switching my Canadian equity holding from plain vanilla to HXT. However with my low income, the tax on Canadian dividends is pretty trivial, so I thought I’d pay a small amount now rather than an undetermined amount on accumulated realised capital gains on the VCE and VCN later (likely on my death, with a huge amount of income that year). One might make that same argument for reintroducing some ZPR in my taxable portfolio, or even BXF again, rather than putting all my bond component in HBB although that last one is not dividend income.
I am sorry but I still don’t see why you would not include a preferred ETF in your RRSP or TFSA as a stream of fixed income. The rates of all the other fixed income ETFs or GICs will leave me eating cat food. Can someone please explain this to me?
@Lindsey: It’s important to consider the bigger picture. The goal should be total return for the least amount of risk, not simply a stream of fixed income. It’s true that bonds and GICs are not likely to deliver high enough returns for most investors to meet their goals these days. But the solution is to build a diversified portfolio that also includes equities to increase those expected returns (albeit with more risk). While preferred shares have higher yields than bonds and GICs, the extra return is not likely to fully compensate you for the additional risk.
@CCP: Regarding the bigger picture, in addition to the rather low correlation with other assets, we have been told that the risk of Preferred ETFs is higher than that of Bond ETFs. But what is the very long term likelihood of total return? Does it fall into the same general range of, say, Canadian Equities?
@oldie, your answer is in Table 4 (10 year vs. risk comparisons) in Raymond and Dan’s paper. In summary the pre-tax return is lower than short term bonds and the risk is double.
@BrianG: Thanks, I should have seen that.
“In summary the pre-tax return is lower than short term bonds and the risk is double.” In a nutshell, this explains why its addition would not be an enhancement of the Plain Vanilla CCP portfolio, or even a tweaked one (with REIT) as demonstrated in Table 5.
Aleady have an allocation for REITs, and laddered GICs. I was just looking for something in the registered accounts that would provide some income for the F.I. side. What is the risk in having 3 % of our registered investments in preferred ETFs?
@CCP, ah, too bad. ;) And to clarify, I expect the answer is “yes” that active management can theoretically provide value in an illiquid market like preferreds, not “yes” that you must have interviewed him. (Although I would recommend reaching out; I bet he’d be good for a quote at least!)
In Chart 2 ($100 Invested in Preferred Shares) and Chart 3 ($100 Invested in Preferred Shares, Stocks and Bonds) in the White Paper referred to at the end of the article, are the returns total returns (i.e. including all dividends)?
@Nathan @ CCP:
I agree with Nathan, an interview with Mr James Hymas would be very interesting following your posts on preferred shares.
@CCP: Given that keeping things simple is very important for DIY investors, I like to think of everything in investing as either a stock or a bond. We own stocks for growth and inflation protection, and bonds for reducing volatility to help us stay the course. To add in a hybrid security like preferred shares, that are neither stocks or bonds, seems to be adding unnecessary complexity.
CCP model portfolios (current) are great for accumulation phase investors but in retirement many use preferred shares in both registered/unregistered accounts. What needs to be discussed is sequence of returns risk which is only applicable in the deaccumulation phase. If you retire and equities plunge and continue down for a number of years, you run the risk of dying poor! This is why if you choose retirement income based on yield variance of portfolio principal value is important for retirees only to the extent that net cash withdrawals exist. Since income is generally much more stable than principal value – for equities as well as for bonds retirees who invest in bonds/GICs, dividends, REITS, preferred shares etc. are less worried if market conditions deteriorate.
@Edwin: They’re total returns. Unless otherwise specified, it’s safe to assume that published returns include reinvested dividends or interest. That is the only meaningful way to measure an investment’s return.
@harveym: Using income-oriented investments does not protect you from sequence of returns risk. I wrote about this recently in MoneySense and I am considering repackaging these ideas in a series of blog posts if readers are interested. In my opinion one of the most misunderstood ideas in investing is that in retirement you need to change your strategy to focus on yield:
http://www.moneysense.ca/retire/selling-etfs-to-generate-retirement-income
Just to complicate things a little more. What about using corporate class bonds relative to preferreds and other bonds in a taxable account? That way we can get full bond exposure and limit the taxes as well….
OR….back to my favorite. HBB and limit taxes altogether until you sell and get compounding in the meantime…..that still seems the most tax efficient out of everything available right now. is that not correct?
@CCP: I’d be very interested in the analysis about changing the portfolio makeup in retirement.
Intuitively, going for stable, income-producing but low-tax products would *seem* like a good idea, but the numbers might work out differently when you do the math and consider all angles.
@Bruce: The tax benefit of corporate class funds, in my opinion, is oversold. The best you can hope for is a deferral of capital gains: they don’t solve the problem of highly taxed bond coupons. And in many cases, the additional fees outweigh any tax benefit.
@CCP It depends what you mean by “focusing on yield”. Chasing yield is fraught with risk and is to be avoided, but for many retirees income investing based on req’d yield may be realistic and easier than your example in Moneysense of total return investment. In fact, the complexity in that article makes the case that you need an advisor since a DIY investor couldn’t manage it.
A retired couple at 65 who had together amassed one million in savings could retire comfortably with joint CCP/OAS plus income from their investment portfolio at an easily attainable rate of return of 3% by investing in a combination of ZPL, CPD, and XEI etfs. This would give them a comfortable retirement with the advantages of easy annual rebalancing to their chosen asset allocations, stable annual income, and less worry regarding portfolio principal variance. Is this not a workable approach?
@harveym: It’s not my intention to talk anyone out of whatever they happen to be doing. My goal is to counter the very common misconception that accumulators and retirees need to use different funds and/or fundamentally different approaches. There are definitely changes in the way a portfolio needs to be managed to generate cash flow, but this has nothing to do with changing investment strategies (i.e. focusing on dividend stocks, using preferred shares if you didn’t use them before, buying REITs for income, etc.). The MoneySense article will probably address all of your concerns, but I will put together something further on this topic.
@CCP Very interested in a blog series about generating retirement income. Perhaps a look at the role that annuities can play in this as well? Thanks for the ongoing education…much appreciated.
@CCP “but I will put together something further on this topic”
Thank-you. Well there are no holy grails in asset management and strategies come and go. Admittedly the arguments in favor of a total return focus are holding sway in the money management industry for the time being. That particular strategy as I’ve said is more complex in retirement and so the industry interest is understandable!
I think that your word “misconception” is unfortunate since it has the connotation of saying mistake which I don’t think you mean. Income investing in retirement is not going away and there are good reasons for that.
http://portfolioist.com/2012/12/10/investing-for-income-vs-total-return/
http://www.himivest.com/media/SecurityIncome.pdf
@harveym, all that matters for a rational investor is return vs. risk. Retirees should determine their realistic target return and then structure a portfolio that allows them to achieve that return at the lowest risk. Yes, this might require an investment in time and learning and/or paying for fee based expert to give you that advice but that’s what is retired to avoid making bad decisions.
I am living off of my investments and I took the DIY globally diversified total return approach (after investigating all approaches) because I determined it has lower risk for a given target return. I won’t go into all the math and analysis I did to come to this conclusion, but it is possible to work out. If math isn’t your thing, a good paid advisor with good software run the numbers. and generate some graphs to demonstrate this.
Also, the total return approach doesn’t have to be complicated. For instance, Tangerine Balanced Income Fund is dead simple; is passive and the equity portion is globally diversified. It would work well for some people where the equity/fixed income mix is what they need. If you wanted to walk on the wild side of active investments (which hasn’t been show to be irrelevant for fixed income investing), Steadyhand Income fund could be another choice.
@harveyM
I think the second article you linked make a few good points. In particular, that once you’re retired, your investment time horizon shortens dramatically and you are left with very little time for investments to recover after big market drops. As a consequence, after you retire your investments should reflect your reduced tolerance for this risk. Generally, this means investing in bonds, GICs, and (maybe) lower risk stocks.
However, I dislike his characterization of this as “Income Investing”. I think that’s very misleading, because you should really be focusing on “total return” rather than “income”, again where “total return” should be subject to your risk profile.
The first article you linked is of low quality. He correctly states that dividend paying companies are less risky, because they are have generally boring business models. He then gets confused when talking about income and returns. The following statement in particular is just wrong:
“We can predict quite confidently that we know how much gain we will get from dividends but we cannot predict the changes in price for AEP over the next year with any confidence at all.”
In fact, you know that (all else equal) the price will go DOWN to offset the amount of dividend paid. The total return from the dividend paying stock should be the same regardless of whether it pays dividends, or reinvests all cash the business produces.
The fact it pays dividends is only useful as an indicator that it’s a “safer” company, and may be a good fit for the growth part of a lower-risk portfolio. The focus on “income investing” is irrelevant.
@Malnar wrote: “your investment time horizon shortens dramatically and you are left with very little time for investments to recover after big market drops.”
I faced this and determined that it is not really true for most people in good health. If you retire at 65, you still have an average life expectancy of 85; up to a 20 year time horizon. A 20 year investment horizon is a fine candidate for balanced portfolio. :) If you retire earlier than that, it only get’s longer.
Of course you’ll say, what about Sequence of Return risk? A simple way to deal with this is to think of your investment in terms of 20 buckets, each labelled with the year when you’ll need it. E.g. bucket 1 is what you’ll need in year 1, bucket 2 is what you’ll need in year 2, etc.
Now invest each bucket appropriately based on the time horizon of when you’ll need it. E.g. buckets 1-5 could be in a 1-5 year GIC ladder, buckets 6-10 in an income fund with an appropriate time horizon and buckets 10-20 in a balanced portfolio. (This is hypothetical, you might have a better idea of how to invest each bucket but just make sure each investment is appropriate for time horizon,)
A neat thing happens when you rebalance this each year. If everything is going well, you can just move one bucket from income into the GIC ladder and one bucket from the balance portfolio into the income portfolio. In the worst case scenario, if the balanced fund tanks, you have a choice… you could choose to do nothing and wait it out. In fact you could wait for up to 10 years for the stocks to recover before you rebalance. Longer if you reduce your spending.
This approach would have served a retiree very well through 2008 for instance. From the data I’ve seen, it would have even worked through the 1929 crash so it seems robust.
Oh… one other addition to the above strategy. If after 10 years, at age 75 your are worried about outliving your funds, buy an annuity. You get much better annuity payouts when you are older because the insurance companies know the average life span is about 85 years. Thus an annuity at this age becomes a pure insurance product (and not an investment product), which is what they are good for.
Thank you all for your comments, they have been very helpful. I particularly like the bucket analogy.
@CCP
http://www.moneysense.ca/retire/selling-etfs-to-generate-retirement-income
Great article! My proposed approach to cash flow when I retire is the same as the article’s, and it really demonstrates it in a clear manner with the charts of asset allocation each year. Thanks.
@All
This is Dan’s blog and I respect his work and his passion for total return investing. It just doesn’t work for me in retirement but I’m glad for those who have made it a success.
In the Moneysense article above Dan does say: “Granted, it would be ideal to live off just the dividends and interest from your portfolio, because if you never have to eat into your capital by selling investments, you’ll never run out of money.” This strategy he says only works for the “wealthy” and I’m certainly not that. However, it can work with a million or less and I believe it does mitigate longevity risk as well as sequence of return risk as illustrated in the second article above. I look forward to Dan’s articles on total return deaccumulation phase strategies.
@harveyM. There is a whole lot of math and advanced concepts in your James Hymas article and I’m not going to pretend I understand it. However, I think I can identify the difference of opinion between you and CCP.
In the opening paragraphs of his article Hymas identifies Security of Principal and Security of Income as opposing outcomes. “…the more you have of one, the less you have of the other” he writes. He makes his point in Table SI-4 — once you have achieved Security of Income you no longer need to care about portfolio performance, cash flow, etc. In this day and age of seniors outliving their savings, a Security of Income approach does make a lot of sense.
However, as CCP has pointed out, the asset classes you would pursue to achieve Security of Income have their downsides. As Hymas has pointed out, you lose Security of Principal by pursuing Security of Income. By trying to secure your Income you’re risking your Principal.
If your primary objective is to not run out of cash flow during retirement and you don’t care how much Principal is left behind then this would be a good strategy. Hymas himself writes “Market interest rates may go up or down; the capital value of his portfolio may change dramatically; but he can blithely cash his government cheques.”
As great as this sounds, there is a common scenario where people do care about the principal left in their portfolio when they die: inheritance. For people wishing to leave behind a sizable estate a Security of Income approach would not be the best strategy.
This is a great topic and I, too, look forward to posts on portfolio withdrawal strategies. I want my portfolio to be both inheritance-preserving and retirement-supporting. Is this asking for too much for people without large portfolios?
Despite my lack of enthusiasm for them, I’ve been reading up more on preferred shares. One thing I came across that I’ve never thought about is that most preferred shares are callable (by the company). If a company calls a preferred share I suspect that means a realized capital gain (or loss) for you holding it. Wouldn’t this throw a big monkey wrench in tax planning? Thus negating some of their supposed tax benefits? Or am I missing something?
Also, from this report, http://www.raymondjames.ca/lewisrosen/preferred.aspx
They state “Of the C$10 bln with resets in 2014, just over 80% or $8 bln were redeemed.”
So it appears not to be just a hypothetical.
Am I correct or missing something?
I just don’t understand why the CPD (S&P/TSX CANADIAN PREFERRED SHARE INDEX ETF) that I bought in May 2011 for $17.40 has dropped recently to $15.40. How much further can it drop?
It’s been generating dividends – consistently – of 4.8%. This doesn’t cover the 11% overall drop, though!
@Alan, James Hymas provides a bit of context on the recent downdraft in FixedResets (which make up a large part of CPD) in the comment section of his Jan 29 post: http://prefblog.com/?p=27622#comment-193144. Worth a read. My interpretation of his explanation is that the market is waking up to the realization that FixedResets are perpetual, and the fact that they can be reset at rates lower than their issue is slapping a few people in the face.
@CCP – thank you for the series. It is timely for me, as I have been recently trying to understand the complexities of preferred shares. Keep up the great work, your blog is an incredible resource for DIY investors!
Another vote for some further discussion of generating income in retirement. Or perhaps, I just need further education in other areas. I just find it mind-boggling to commit upwards of 50% of my capital to bond ETFs at this juncture in accordance with a cautious approach to the equities market. I’ve read the articles you’ve mentioned previously with regard to bonds, but I feel somewhat like a deer in the headlights…frozen…
Thanks again for this wonderful blog and your time and effort in maintaining same.
@Ly: Thanks for the kind words. I can certainly understand the analysis paralysis. But you may want to think about getting help from an unbiased financial planner who can help set you on the right track.
@harveyM I would say that the main reason you don’t risk running out of money if you live off interest and dividends from a balanced portfolio isn’t because you’re not dipping into capital, but more fundamentally, because you must only be withdrawing 2% or so. If, on the other hand, you structure a portfolio that spits out 6% in interest and dividends, you take at least the same risk of running out of money as someone withdrawing 6% a year from the first portfolio. Distributions aren’t guaranteed forever. Dividends can be cut. Bonds (especially high-yield ones) can default. Those things may not be likely, but they’re likely enough that your risk of ruin there is at least as great as in the “dipping into capital” case.
@Nathan
In retirement your money worries change and are focused on annual cash flow along with sequence of returns risk and longevity risk. Annual portfolio withdrawl must be dynamic based upon a realistic budget which will not remain static unfortunately. In retirement there is a difference between investing with the hope of asset appreciation and investing on the basis of income. The former is anxiety provoking while the latter less because your estimation risk for future return is lower with income investing than for a non-income generating asset. For example, I can be more certain that a portfolio of ZPL+CPD+XEI+ZRE will give me at least 3% return for the next twelve months whereas a portfolio of VCN+XAW may not!
@Brian G
I just wanted to correct your life expectancy stats, since your rule of thumb is a bit outdated.
A 65 year old (in 2015) will live on average to age 87 or 90, for males and females respectively.
A 75 year old (in 2015) will live on average to age 89 or 91, for males and females respectively.
Generally, the longer you live, the better chance that you’ll live even longer.
My stats come from a recent (2014) Canadian Institute of Actuaries pensioner mortality study.
@JRS Many thanks for your information, I missed that news and now understand what happened. Even though I had done some searching several times on the internet to try to discover what had happened. These are the downside of the DIY investor, I suppose.
Thanks again.
Hello,
Since the beginning of the year, preferred ETFs have decreased significantly. Ie: CPD fomr the mid 16$ to the 14$ range. Yield is now around 5%.
Is this explained? And, is it a good time to increase one’s position?
Thoughts?
Eric
@Eric: This may be helpful:
https://www.pwlcapital.com/en/The-Firm/Research-span-Department-span-/Blog/Research-Blog/April-2015/The-Volatility-of-Fixed-Reset-Preferred-Shares